• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.91-98
• /
• 2023

Recently, according to development of artificial intelligence, a wide range of industry being automatic and optimized. Also we can find out some research of using supervised learning for deteceting defect of railway in domestic rail industry. However, there are structures other than rails on the track, and the fastener is a device that binds the rail to other structures, and periodic inspections are required to prevent safety accidents. In this paper, we present a method of reducing cost for labeling using semi-supervised and transfer model trained on rail fastener data. We use Resnet50 as the backbone network pretrained on ImageNet. At first we randomly take training data from unlabeled data and then labeled that data to train model. After predict unlabeled data by trained model, we adopted a method of adding the data with the highest probability for each class to the training data by a predetermined size. Futhermore, we also conducted some experiments to investigate the influence of the number of initially labeled data. As a result of the experiment, model reaches 92% accuracy which has a performance difference of around 5% compared to supervised learning. This is expected to improve the performance of the classifier by using relatively few labels without additional labeling processes through the proposed method.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.119-125
• /
• 2023

Through the Ukraine-Russia war, the military importance of drones is being reassessed, and North Korea has completed actual verification through a drone provocation towards South Korea at 2022. Furthermore, North Korea is actively integrating artificial intelligence (AI) technology into drones, highlighting the increasing threat posed by drones. In response, the Republic of Korea military has established Drone Operations Command(DOC) and implemented various drone defense systems. However, there is a concern that the efforts to enhance capabilities are disproportionately focused on striking systems, making it challenging to effectively counter swarm drone attacks. Particularly, Air Force bases located adjacent to urban areas face significant limitations in the use of traditional air defense weapons due to concerns about civilian casualties. Therefore, this study proposes a new passive air defense method that aims at disrupting the object detection capabilities of AI models to enhance the survivability of friendly aircraft against the threat posed by AI based swarm drones. Using laser-based adversarial examples, the study seeks to degrade the recognition accuracy of object recognition AI installed on enemy drones. Experimental results using synthetic images and precision-reduced models confirmed that the proposed method decreased the recognition accuracy of object recognition AI, which was initially approximately 95%, to around 0-15% after the application of the proposed method, thereby validating the effectiveness of the proposed method.

• Journal of Internet Computing and Services
• /
• v.24 no.6
• /
• pp.161-169
• /
• 2023

Due to the Fourth Industrial Revolution, innovation in various technologies such as artificial intelligence (AI), big data (Big Data), and cloud (Cloud) is accelerating, and data is considered an important asset. With the innovation of these technologies, various efforts are being made to lead technological innovation in the field of defense science and technology. In Korea, the government also announced the "Defense Innovation 4.0 Plan," which consists of five key points and 16 tasks to foster advanced science and technology forces in March 2023. The plan also includes the establishment of a Condition-Based Maintenance system (CBM+) to improve the operability and availability of weapons systems and reduce defense costs. Condition Based Maintenance (CBM) aims to secure the reliability and availability of the weapon system and analyze changes in equipment's state information to identify them as signs of failure and defects, and CBM+ is a concept that adds Remaining Useful Life prediction technology to the existing CBM concept [1]. In order to establish a CBM+ system for the weapon system, sensors are installed and sensor data are required to obtain condition information of the weapon system. In this paper, we propose a sensor data metadata schema to efficiently and effectively manage sensor data collected from sensors installed in various weapons systems.

• Journal of Internet of Things and Convergence
• /
• v.9 no.6
• /
• pp.57-62
• /
• 2023

This research is a comparative analysis of the U.S. S&P 500 index using the volatility breakout strategy against the Buy and Hold approach. The volatility breakout strategy is a trading method that exploits price movements after periods of relative market stability or concentration. Specifically, it is observed that large price movements tend to occur more frequently after periods of low volatility. When a stock moves within a narrow price range for a while and then suddenly rises or falls, it is expected to continue moving in that direction. To capitalize on these movements, traders adopt the volatility breakout strategy. The 'k' value is used as a multiplier applied to a measure of recent market volatility. One method of measuring volatility is the Average True Range (ATR), which represents the difference between the highest and lowest prices of recent trading days. The 'k' value plays a crucial role for traders in setting their trade threshold. This study calculated the 'k' value at a general level and compared its returns with the Buy and Hold strategy, finding that algorithmic trading using the volatility breakout strategy achieved slightly higher returns. In the future, we plan to present simulation results for maximizing returns by determining the optimal 'k' value for automated trading of the S&P 500 index using artificial intelligence deep learning techniques.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.10
• /
• pp.455-460
• /
• 2023

Text augmentation is a methodology that creates new augmented texts by transforming or generating original texts for the purpose of improving the performance of NLP models. However existing text augmentation techniques have limitations such as lack of expressive diversity semantic distortion and limited number of augmented texts. Recently text augmentation using large language models and few-shot learning can overcome these limitations but there is also a risk of noise generation due to incorrect generation. In this paper, we propose a text augmentation method called TAGS that generates multiple candidate texts and selects the appropriate text as the augmented text. TAGS generates various expressions using few-shot learning while effectively selecting suitable data even with a small amount of original text by using contrastive learning and similarity comparison. We applied this method to task-oriented chatbot data and achieved more than sixty times quantitative improvement. We also analyzed the generated texts to confirm that they produced semantically and expressively diverse texts compared to the original texts. Moreover, we trained and evaluated a classification model using the augmented texts and showed that it improved the performance by more than 0.1915, confirming that it helps to improve the actual model performance.

• Korean Journal of Remote Sensing
• /
• v.39 no.5_4
• /
• pp.1125-1134
• /
• 2023

Out of the total 17,000 reservoirs in Korea, 13,600 small agricultural reservoirs do not have hydrological measurement facilities, making it difficult to predict water storage volume and appropriate operation. This paper examined univariate and multivariate long short-term memory (LSTM) modeling to predict the storage rate of agricultural reservoirs using remote sensing and artificial intelligence. The univariate LSTM model used only water storage rate as an explanatory variable, and the multivariate LSTM model added n-day accumulative precipitation and date of year (DOY) as explanatory variables. They were trained using eight years data (2013 to 2020) for Idong Reservoir, and the predictions of the daily water storage in 2021 were validated for accuracy assessment. The univariate showed the root-mean square error (RMSE) of 1.04%, 2.52%, and 4.18% for the one, three, and five-day predictions. The multivariate model showed the RMSE 0.98%, 1.95%, and 2.76% for the one, three, and five-day predictions. In addition to the time-series storage rate, DOY and daily and 5-day cumulative precipitation variables were more significant than others for the daily model, which means that the temporal range of the impacts of precipitation on the everyday water storage rate was approximately five days.

• Korean Journal of Ecology and Environment
• /
• v.56 no.3
• /
• pp.268-279
• /
• 2023

Nowadays, artificial intelligence model approaches such as machine and deep learning have been widely used to predict variations of water quality in various freshwater bodies. In particular, many researchers have tried to predict the occurrence of cyanobacterial blooms in inland water, which pose a threat to human health and aquatic ecosystems. Therefore, the objective of this study were to: 1) review studies on the application of machine learning models for predicting the occurrence of cyanobacterial blooms and its metabolites and 2) prospect for future study on the prediction of cyanobacteria by machine learning models including deep learning. In this study, a systematic literature search and review were conducted using SCOPUS, which is Elsevier's abstract and citation database. The key results showed that deep learning models were usually used to predict cyanobacterial cells, while machine learning models focused on predicting cyanobacterial metabolites such as concentrations of microcystin, geosmin, and 2-methylisoborneol (2-MIB) in reservoirs. There was a distinct difference in the use of input variables to predict cyanobacterial cells and metabolites. The application of deep learning models through the construction of big data may be encouraged to build accurate models to predict cyanobacterial metabolites.

• Education of Primary School Mathematics
• /
• v.26 no.4
• /
• pp.333-348
• /
• 2023

Clear analysis and diagnosis of various characteristic factors of individual students is the most important in order to realize individual customized teaching and learning, which is considered the most essential function of math artificial intelligence-based digital textbooks. In this study, analysis factors and tools for individual customized learning diagnosis and construction models for data collection and analysis were derived from mathematical AI digital textbooks. To this end, according to the Ministry of Education's recent plan to apply AI digital textbooks, the demand for AI digital textbooks in mathematics, personalized learning and prior research on data for it, and factors for learner analysis in mathematics digital platforms were reviewed. As a result of the study, the researcher summarized the factors for learning analysis as factors for learning readiness, process and performance, achievement, weakness, and propensity analysis as factors for learning duration, problem solving time, concentration, math learning habits, and emotional analysis as factors for confidence, interest, anxiety, learning motivation, value perception, and attitude analysis as factors for learning analysis. In addition, the researcher proposed noon data on the problem, learning progress rate, screen recording data on student activities, event data, eye tracking device, and self-response questionnaires as data collection tools for these factors. Finally, a data collection model was proposed that time-series these factors before, during, and after learning.

• Journal of Advanced Navigation Technology
• /
• v.27 no.4
• /
• pp.447-452
• /
• 2023

In this paper, gain enhancement of a double dipole quasi-Yagi antenna using a meanderline array structure was studied. A 4×1 meanderline array structure consisting of a meanderline conductor- shaped unit cell is located above the second dipole of the double dipole quasi-Yagi antenna. It was designed to have gain over 7 dBi in the frequency range between 1.70 and 2.70 GHz in order to compare the performance with the case using a conventional strip director. As a result of comparison, the average gain of the double dipole quasi-yagi antenna with the proposed meander line array structure was larger compared to the case with the conventional strip director. A double dipole quasi-Yagi antenna using the proposed meanderline array structure was fabricated on an FR4 substrate and its characteristics were compared with the simulation results. Experiment results show that the frequency band for a VSWR less than 2 was 1.55-2.82 GHz, and the frequency band for gain over 7 dBi was measured to be 1.54-2.83 GHz. The frequency bandwidth with gain over 7 dBi increased, and average gain also slightly increased, compared to the conventional case using a strip director.

• Phonetics and Speech Sciences
• /
• v.15 no.4
• /
• pp.71-80
• /
• 2023

This study primarily aimed to develop an automated stuttering identification and classification method using artificial intelligence technology. In particular, this study aimed to develop a deep learning-based identification model utilizing the convolutional neural networks (CNNs) algorithm for Korean speakers who stutter. To this aim, speech data were collected from 9 adults who stutter and 9 normally-fluent speakers. The data were automatically segmented at the phrasal level using Google Cloud speech-to-text (STT), and labels such as 'fluent', 'blockage', prolongation', and 'repetition' were assigned to them. Mel frequency cepstral coefficients (MFCCs) and the CNN-based classifier were also used for detecting and classifying each type of the stuttered disfluency. However, in the case of prolongation, five results were found and, therefore, excluded from the classifier model. Results showed that the accuracy of the CNN classifier was 0.96, and the F1-score for classification performance was as follows: 'fluent' 1.00, 'blockage' 0.67, and 'repetition' 0.74. Although the effectiveness of the automatic classification identifier was validated using CNNs to detect the stuttered disfluencies, the performance was found to be inadequate especially for the blockage and prolongation types. Consequently, the establishment of a big speech database for collecting data based on the types of stuttered disfluencies was identified as a necessary foundation for improving classification performance.